Health is “the real wealth” and Artificial Intelligence (AI) is now used more than ever by healthcare organizations to help improve patient care, reduce costs, increase revenue, and reduce risk.
OMNImed™ SmartOR is just one example of this. OMNImed is a smart operating room that leverages AI, machine learning, Internet of Things (IoT) and computer vision to monitor and document every aspect of a procedure. It can track up to 10 million data points per minute. The goal with this technology is to reduce infections post-surgery.
AI and Machine Learning have been just within reach for a while now. However, with a shortage of data scientists, organizations don’t always know where to get started, can’t get an ROI quick enough, and once their AI/ML is operational, they don’t have the technology to manage and monitor the models. Fortunately, there is a solution to help bridge the gap to these needs.
Here are 6 questions that leaders at healthcare organizations should be asking themselves to help increase revenue, reduce cost and reduce risk within production and planning:
1. How can we help improve our procedures for identifying patients who are most at risk of 30-day readmission?
According to Kaiser Health News, 9 in 10 general acute-care hospitals have been penalized at least once in the past decade for readmissions.
Reporting based on data from the Centers for Medicare and Medicaid Services shows over the lifetime of the program, 2,920 hospitals have been penalized at least once. That’s 93% of the 3,139 general acute hospitals subject to HRRP evaluation, and 55% of all hospitals. Moreover, 1,288 have been punished in all 10 years.
2. Are we accurately predicting length of stay (LOS) for patients at admission stage?
The average insured overnight hospital stay costs about $11,700 and every extra day a patient is at a hospital is a day that a new patient can’t be treated.
There is a need to help improve patient outcome by more accurately predicting a patient’s (length of stay) LOS at the admission stage. According to the Journal of Healthcare Engineering, researchers were correctly able to predict with 88.31% to 91.53% accuracy the LOS at the preadmission stage for patients with coronary atherosclerosis (CAS) by using an artificial neural network.
3. What can we do to help improve medication adherence among our patients in order to reach compliance with value-based care initiatives? Can we identify a patient’s likelihood to admit at the triage stage to adhere with our value-based care policies?
According to a study done by CDC, medication nonadherence is associated with higher rates of hospital admissions, suboptimal health outcomes, increased morbidity and mortality, as well as increased health care costs. In the United States, of the 3.8 billion prescriptions written annually, approximately 20% of the new prescriptions are never filled, and among those filled, approximately 50% are taken incorrectly. Direct health care costs associated with nonadherence have grown to approximately $100–$300 billion of US health care dollars spent annually.
4. Is there a way to forecast incoming inpatient volume at our hospitals to improve staffing on hospital floors?
60% of hospital operational cost comes from staffing. There is a need to help improve patient care by adequately staffing hospital floors by predicting expected incoming patient volume –
Currently most hospitals use historical averages to forecast expected number of incoming patients. Such forecasts are used to staff hospital floors for the coming day/week/month. According to a study done by researchers at Northwestern University, AI/ML forecasting models outperformed the historical average-based approach by reducing Mean Absolute Percentage Error (MAPE) from 17.2% to 6% in one day ahead forecast and to 8.8% MAPE in a month ahead forecast.
5. During which days/times should we double or triple book our outpatient appointments to minimize downtime for our doctors?
There is a need to reduce staff downtime by predicting which outpatient appointments will be cancelled – According to a study published by National Institute of Health, on average, 18.8% of the outpatient appointments are cancelled at the last minute. The cancellation rate was as high as 27% in some cases. According to the same study each cancelled appointment cost the provider $254.